Developing and testing seed mixes for pollinators

As agricultural intensification and urbanisation increases, gardens and amenity spaces are becoming increasingly important habitats to support pollinators. This is vital, since our pollinators have declined considerably in the last 50 years and are critically important for the pollination of both agriculturally important crops and flowering plants in the wider ecosystem.

‘Plants for pollinators’ lists can be useful to help us choose the right plants, but they are generally not based on real scientific data and miss out important plants that support pollinators.

PhD student Lucy Witter has been investigating commercially available seed mixes to find out if they can provide sufficient resources for a diversity of wild pollinators (bumblebees, hoverflies and solitary bees). At the Garden there are two sites at which trial plots have been sown, these consist of four different seed mixes: two marketed as “pollinator friendly”, one native mix and one designed for aesthetic purposes. These are being studied using observational methods and DNA metabarcoding techniques to determine what proportion of species from the pollinator friendly seed mixes are being used compared to the resources in the surrounding environment.

As well as this research, Lucy has undertaken a systematic review of published literature on plant-pollinator interactions and used this information to develop two experimental seed mixes designed to attract a diversity of wild pollinator species. Two experimental seed mixes were trialed (one native and one non-native) at the sites previously mentioned and compared to commercially available seed mixes using observational surveys. Visitors were also asked on their opinion on the attractiveness of the seed mixes, in order to compare public preference to pollinator choice.


Witter L. (2021) Plants for Pollinators: Assessing the ability of “pollinator friendly” annual seed mixtures designed of gardens and amenity space to support wild pollinators PhD thesis